Late to the party but didn't really expected that sbmm is so shallow

We all saw what was datamined but I really thought there was more to it.
Not that I didn't thought escapes and kills were most important but what happened to "match time" etc?
If it is only kills and escapes that is even more simple than emblems and ranks. :S
I was fine with sbmm but if it is this simple it is not good at all considering we are talking about dbd here.
Comments
-
I still believe time is a factor. Not that it really makes any difference in the grand scheme of things.
2 -
True but I wish there was more to sbmm tbh.
I was usually altruistic survivor and it made me less of that so I can get "better" teammates.
1k raising mmr is even more of a blow to me honestly. How can you consider 1k a win?
2 -
I don't think it is. I think it's basically going to look at kills vs. escapes. For example if every kill is plus 50 points and each escape is a minus 50 a 1k you'd lose mmr. I think that's the basic gist for killer. Survivor seems to be if you didn't escape you're trash.
1 -
Oh ok I saw some people saying 1k is a win so I was concerned.
I think 3k should raise mmr but nothing lower than that.
0 -
Sbmm is arguably worse then the old emblem system in every way and with its "soft caps" will only get worse over time even with bhvr adjustments
8 -
Because the dev never said how many kill is a win for the killer that why some people say or think 1 kill is a win
1 -
As far as I understand that's basically how it works. There's supposedly some more complicated math involved, but that's essentially how it works.
1 -
yes it is sadly shallow, still feel like more factors should be added to make it a full system but the survivors losing less if they are the 2nd 3rd or 4th one to die is actually a pretty good factor I wouldn't have thought of
2 -
Time acts kind of as a moderator from what I saw, so if you are playing killer and got a 4k in 5 minutes you get like 30pts but if it was in 10 minutes you would get 40pts. Though again numbers are just examples we don't know the exact math to my knowledge.
I think sbmm is basically the same as the emblem system for killers, since killers who get kills tend to do better in emblems and meet the goals of sbmm, but survivors get screwed. Good survivors who risk their lives to save teammates will be considered worse then the teammates they rescued, and the guy sitting in the corner all game with a key is going to end up looking better then all of his teammates if he is the one who survives.
Honestly the system is way to basic for something that took over a year to develop and is largely the same or worse then the emblem system, especially with the hard cap.
3 -
Good survivors who risk their lives to save teammates will be considered worse then the teammates they rescued, and the guy sitting in the corner all game with a key is going to end up looking better then all of his teammates if he is the one who survives.
I think that's a bit of a wrong conception we are having. Largelly because of how poorly sbmm was named.
Having higher mmr does not mean you are better. It's not trying to put a number on how good you are.
What it's trying to do is to get you to a 50% escape sweetspot.
So if you like to make risky resques your mmr will change so that you can make risky resques AND escape about half the time.
It's not looking at how good you are, it's looking so you can escape half the time no matter what your playstyle is
It's just a wrong conception people are having that high mmr equals you are good and low mmr equals you are bad.
It's just a number that tries to give you a 50% escape rate no matter what your playstyle is
1 -
urban evasion now meta
2 -
It is a bit more shallow than I hoped, but after thinking about it, it actually makes a lot of sense for me.
Effectively, there are 2 systems at play: MMR for balance, and grading for fairnesa.
Balance is to make sure people have a 50/50 chance of doing well, and therefore try to give everyone a chance of "winning". A lot of people would be gutted if they never escaped or consistently got 0-1k. This system allows that opportunity and therefore hopes to balance the game for those players. Also, those who escape a lot will find tougher challenges of escaping.
However, balance is not the same as fairness, because balance doesn't care how people get to a certain point; only that you gain a balanced experience. Fairness is about the effort and time you put in during the match, which bring us to the grading:
Grading is about everything you do during the trial and time put into the game. This rewards players by utilising the Emblem system to gain higher grades and end-of-month rewards. This is the reward section of the game, but is separate from winning/ losing, because you can get a lucky escape through the gates by doing almost nothing and working selfishly, or getting a 4k solely through NOED.
Overall, the system makes a lot of sense. I'm not sure that the PR behind it was well done, and the lead designer - whilst I understand what he was saying about "wouldn't it be great to do those things and win" bit, and that it wasn't mocking but moreover saying MMR would allow people who put the effort in but are ultimately unlucky to get better chances in future games - did present this message in a way that could be misinterpreted as mocking.
There's a need to understand ranking no longer exists, and this system is about increasing the balance and fairness of the game in the best way possible, especially in a game system that effectively can't be truely balanced or fair.
1 -
That may be true but that is still just as faulty as the old system, since the old system was largely based on the same criteria, and assumes everyone always plays roughly the same all the time. A better system could account for things like perks to place you in stronger matches when you run better stuff or weaker with weaker stuff. Same is true for if someone is playing with friends or not or if you bring really strong add-ons as killer.
The point of the system should be to make matches fair and fun (like all sbmm systems) and when you only look at one criteria you miss to much of what is going on around that criteria.
1 -
Idk, the old system really screwed certain killers (poor plague)
I think there is just too much criteria to make a reliable system that we were expecting.
Take your example that the system looks at perks. That would mean every 3 months 6 new perks need to be added to be calculated. Not to mention any synergies that might make other perks stronger.
That would be an insane amount of work.
What we have now is flawed. But i don't think it's possible to have a good system with as many variables dbd has.
1 -
I understood it as when you play killer the MMR is basically a 1v1 between 4 survivors. So if you killed survivor, you would lose against 3 others, meaning your MMR would drop.
2 -
I agree there are too many factors. However, they could make a sub-system that makes this easier, such as a point system for perks. Then the new perks, add-ons, etc could be added to that system and the overall system would not need to change.
Depending on how they did it, a point system that caps how many good perks a person can use could also change the meta and make perk changes easier, while allowing perks that serve the same purpose to actually both have a place in the game. I have talked before about how a system like this could be used to improve balance between perks, different killers, and even swfs, and having something like this in place would then allow for a better SBMM system.
With the current design of the game it is largely impossible as you said. However, I do believe in two years they could have at least made a system that takes more into account then the current one does.
1 -
The most simple solutions are the best solutions.
Killers see kills as wins and Survivors see escapes as wins. Killers use all of their skills to make kills and Survivors use all of their skills to make escapes. The emblem system sucked because it tried to use way to many data points to do the exact same thing as this current system does.
And this system works. Most people like the new system and feel it is more fair.
1 -
Actually, no.
Solutions that are the first thing you think of, and looks sensible, and are easy to implement.
Are often, terrible, ineffective solutions, once implemented, will drag on everyone.
Forever.
cough Old Mettle of Man cough
cough Old Legion changes cough
2 -
Satisfaction surveys disagree with your assessment. I have yet to see a social media outlet which doesn't overwhelmingly describe SBMM as negative.
2 -
Simple solutions for complex problems doesn't mean the solution is the first thing you think of. Usually simple solutions are the hardest to come to because you are eliminating all of the excess complexity that is bogging down the system. Simple solutions are the best because the are the most efficient and fastest solution to a problem without waste.
Think of programming. The first solution you come up with is usually extremely convoluted and wastes processing power. But if you go back to that program, you can simplify it and remove all of "junk" that is unnecessarily wasting time. I've had code that I've went back to and reduced the size of it by half. Simple and eloquent.
Look at the emblem system, it was a bloated piece of crap with all types of variables. There was just no way it would work well with all the useless information going into to determine a fair matchmaking system. The developers went back and pared it down and judging by the community, most people are satisfied with it, as a whole.
1 -
From what I understood from the stream, 1k is a win in the sense that the killer's MMR is calculated from the perspective of them playing 4 games at once. So getting a 4k would be getting 4 wins. What constitutes "winning the trial" for Killer was left undefined, but the goal is that killers should average around 2 kills per trial.
1 -
I love that basically you can farm mmr if you facecamp with bubba and get heavily punished if you get facecamped or tunneled to death as survivor.
I'm an average survivor that has been getting tunneled very frequently so I guess I should be playing with new survivors because of that.
I hated the idea of sbmm before it was introduced and now I hate it even more.
3